The purpose of this project is to extend the efforts of a community- based research consortium established in 1988 to evaluate the effectiveness of culturally appropriate intervention in reducing the spread of AIDS among individuals who engage in high risk behaviors. The specific aims of the proposed project are to test the comparative effectiveness of unenhanced and enhanced prevention programs and to separately test culturally appropriate against non-culturally specific prevention programs in reducing AIDS risk among IDUs, their sex partners, sex workers, and non-IDU cocaine users. Findings from the earlier study (COPE I) tend to support the hypothesis that culturally targeted intervention is potentially is potentially more effective in reducing AIDS risk among ethnic minorities than standard interventions. Project COPE II will draw on experience gained in COPE I to improve the cultural fit of both our African American and Latino culturally appropriate interventions and fortify the standard enhanced model. The project will also compare AIDS prevention findings from these fortified models against outcomes from the original culturally appropriate models and will test the effectiveness of recurrent prevention reinforcement in reducing AIDS risk in groups that tend to relapse into risk behavior after initial intervention. Pre/post-test data for all subjects will be compared and paired comparisons will be made between: 1) the unenhanced and enhanced interventions; 2) the culturally targeted and non-culturally specific standard interventions; 3) the culturally specific intervention in COPE I (based on our existing data set) and the improved culturally specific interventions in COPE II; and 4) the control group (which will receive only initial intervention at intake) and the experimental group (which will receive first and second level enhanced intervention over the course of several months). The project will recruit and assign a sample of 2,500 participants for educational counseling over five years, controlling for ethnicity and target group in random assignment to interventions. These groups will be randomly assigned to intervention so that 20% of the sample (the control group; N=500) will receive unenhanced intervention and 80% will receive enhanced (the experimental group; N=2,000). The latter group will be further assigned on a random basis to culturally appropriate or standard enhanced intervention, so that 60% each of African Americans and Latinos will be assigned to their respective culturally appropriate interventions (N=530 of each), and 40% of these ethnic groups (N-358 of each) and all other ethnic groups (total N=940) will be assigned to the standard model. Process evaluation will describe all interventions as well as further differentiating culturally appropriate from standard intervention content and methods. Outcome measures will include changes in seropositivity, AIDS knowledge, risky sexual practices, needle using/sharing patterns, program participation, morbidity, and mortality.